The thermal history of human fossils and the likelihood of successful DNA amplification

Journal of Human Evolution 45 (2003) 203–217 The thermal history of human fossils and the likelihood of successful DNA amplification Colin I. Smith a1, Andrew T. Chamberlain b, Michael S. Riley c, Chris Stringer d, Matthew J. Collins a2* a Fossil Fuels and Environmental Geochemistry, (Postgraduate Institute); NRG, Drummond Building, University of Newcastle, Newcastle upon Tyne NE1 7RU, UK b Dept of Archaeology and Prehistory, University of Sheffield, Northgate House, Sheffield, S1 4ET, UK c School of Geography, Earth and Environmental Sciences, University of Birmingham, Edgbaston, Birmingham, B15 2TT, UK d Dept of Palaeontology, The Natural History Museum, London, SW7 5BD, UK Received 2 December 2002; accepted 28 July 2003 Abstract Recent success in the amplification of ancient DNA (aDNA) from fossil humans has led to calls for further tests to be carried out on similar material. However, there has been little systematic research on the survival of DNA in the fossil record, even though the environment of the fossil is known to be of paramount importance for the survival of biomolecules over archaeological and geological timescales. A better understanding of aDNA survival would enable research to focus on material with greater chances of successful amplification, thus preventing the unnecessary loss of material and valuable researcher time. We argue that the thermal history of a fossil is a key parameter for the survival of biomolecules. The thermal history of a number of northwest European Neanderthal cave sites is reconstructed here and they are ranked in terms of the relative likelihood of aDNA survival at the sites, under the assumption that DNA depurination is the principal mechanism of degradation. The claims of aDNA amplification from material found at Lake Mungo, Australia, are also considered in the light of the thermal history of this site.  2003 Elsevier Ltd. All rights reserved. Keywords: Ancient DNA; Neanderthal; Modern Humans; Lake Mungo; Thermal History 1 Present address: Museo Nacional de Ciencias Naturales, C/Jose Gutierrez Abascal, 2, 28006, Madrid, Spain. Present address: BioArch, The King’s Manor, York, Y01 7EP. * Corresponding author. Tel.: +44-190-443-3901; fax: +44-191-222-5431. E-mail addresses: csmith@mncn.csic.es (C.I. Smith), A.Chamberlain@Sheffield.ac.uk (A.T. Chamberlain), m.riley@bham.ac.uk (M.S. Riley), C.Stringer@nhm.ac.uk (C. Stringer), mc80@york.ac.uk (M.J. Collins). 2 0047-2484/03/$ - see front matter  2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0047-2484(03)00106-4 204 C.I. Smith et al. / Journal of Human Evolution 45 (2003) 203–217 Introduction The phylogenetic relationships between modern humans (Homo sapiens) and other species of Homo are a source of much debate (e.g. Mellars, 1999). A recent approach to collecting data to investigate the relationships is to isolate genetic material from fossil specimens (e.g. Krings et al., 1997; Ovchinnikov et al., 2000; Krings et al., 2000; Schmitz et al., 2002 for Neanderthals; Adcock et al., 2001 for early Australians), and compare this with modern human DNA. Initial successes have led to calls for further genetic sequences, from different geographic locations and age ranges (Höss, 2000; Adcock et al., 2001); however, the quality of aDNA studies in general, has recently been brought into question (Cooper & Poinar, 2000). The isolation of DNA from fossils is a destructive process, and when dealing with valuable material such as fossil humans it is important that a precautionary approach is taken. Assessment of the likelihood of a sample containing amplifiable DNA should be a prerequisite for such work, maximising returns from researcher time, funding, and most importantly, the valuable and irreplaceable resource of the fossils themselves. It has been observed that DNA appears to survive best in cold dry environments such as permafrost, or high altitude caves, and biochemical studies suggest it is unlikely to survive for more than 100 000 years (Wayne et al., 1999). Yet, there has been little systematic work on the long-term survival of DNA in the fossil environment, and consequently assessments of the prospects of DNA survival remain anecdotal. The need to find further fossil human DNA sequences from wider geographical and temporal ranges is compelling. The proposed upper bound to survival of 100 000 years encompasses both the extinction of the Neanderthals, and the diversification of modern humans, but the oldest successful amplifications are from permafrost, not sites of human occupation. While it is apparent that fossils from cold environments will have better biomolecular preservation than those from hot climates, and that younger fossils will be better preserved than older ones, the distinction between an old and cold fossil and a young hot one is more difficult to assess. Attempts have been made to relate temperature dependent rates of DNA depurination to absolute copy numbers (Pääbo & Wilson, 1991; Marota et al., 2002), but they appear to have been over simplistic and as a result inaccurate (see discussion). The preservation of biomolecules in the fossil environment is complex, especially in bone (Collins et al., 2002). In brief, bone degradation is considered to occur mainly by two processes; one rapid, mediated by microorganisms and fungi (Hackett, 1981; Bell et al., 1996), and the other, chemical degradation, which is a relatively slow process. For skeletal material to become part of the fossil record it is likely that microbial attack will have to be excluded (Trueman & Martill, 2002). If microbial taphonomy is inhibited, then the two major chemical pathways that will lead to DNA destruction are condensation (e.g. glycation of nucleobases, Pischetsrieder et al., 1999) and hydrolysis of the purine bases (Lindahl & Nyberg, 1972). The importance of cross-linking in the survival of DNA has not been investigated in detail so far (see Poinar, 1999). Principal factors that influence the rate of hydrolytic depurination are pH, amount of chemically available water and temperature. The first two factors are less significant in bone as bone itself exerts a substantial buffering effect between pH 4–9 (Bada & Shou, 1980), and the pore size distribution of bone encourages water retention (Hedges & Millard, 1995; Turner-Walker et al., in press). Deep burial will buffer temperature fluctuation, but only around an annual mean, and thus temperature is likely to play a substantial role in defining the envelope of DNA survival (Smith et al., 2001). Here we present a more detailed account of our assessment of the thermal history of fossil hominid sites from Northern Europe, and use this to rank sites according to their thermal age. We define thermal age as the time taken to produce a given degree of DNA degradation when temperature is held at a constant 10(C. The thermal age adjusts the chronological age of different sites according to their individual thermal histories, using the known temperature dependence of DNA depurination estimated in aqueous solution. A comparison is made between the DNA C.I. Smith et al. / Journal of Human Evolution 45 (2003) 203–217 205 depurination thermal ages of sites in NW Europe and Lake Mungo, a site in Australia where controversial claims have been made for the recovery of ancient DNA (Adcock et al., 2001; Cooper et al., 2001). We have attempted to reconstruct the thermal history of Northwest European Neanderthal cave sites, and that of the early anatomically modern human site of Lake Mungo in Australia, by combining both modern day temperature data, and palaeoclimatic evidence. Methods Thermal model for Northwest European Neanderthal cave sites The thermal regime of a fossil is governed by two major factors, the mean temperature and the variation about this mean, both of which will vary over time due to climatic changes. Thus, to reconstruct the thermal history of a fossil, data must be obtained for both the modern day temperature of the site and the palaeotemperature. For a fossil buried in open ground the thermal regime of the fossil is assumed to be the same as that of the surrounding soil. It has been demonstrated that good estimates (1(C) of mean soil temperature can be taken from mean air temperature (MAT) data obtained from local weather stations (Kusada & Achenbach, 1965; Toy et al., 1978), although the thermal regime (i.e. extent of seasonal fluctuation in the soil) is difficult to predict. The amount of seasonal fluctuation in the burial environment, and hence the effective MAT for a chemical reaction (see Wehmiller et al., 2000) will be controlled by many factors including; burial depth, the fluctuation in the climate, the thermal properties of the burial environment (Krarti et al., 1995) and local effects (e.g. vegetation or snow cover). Consequently, assessment of the thermal history of open sites is difficult as such information is not usually available nor can it be readily extrapolated back into the past. In the cave environment the MAT is generally considered a good approximation of local deep cave air temperatures (Bogli, 1980). Deep cave temperatures are also known to be quite static and prone to little seasonal variation, also humidity in caves is high, and not prone to large fluctuations. For a fossil buried in a cave sediment, any fluctuations around the mean value will be dampened further by the cave sediment. It should be considered however that there is more fluctuation at the cave entrance in both temperature and humidity. Modern day mean annual temperatures of the Neanderthal cave sites listed (Table 1) have been estimated from weather station data nearest to each locality. For simplicity, and due to a lack of quantitative palaeoclimate data, the Holocene is considered to be a stable climatic period with constant temperature back to 10.4 kyr BP. Temperatures for this period were estimated from sequences of continuous data (years-decades) from a weather station in the locality of each Neanderthal site (Baker et al., 1994). Only mean temperatures are used in this model, as there is assumed to be no variation about the mean for the burial environment. Altitudes of the cave sites in metres have been estimated from GTOPO30 (Gesch et al., 1999) using a weighted average of the four nearest values to the location unless the correct altitude is known (italics, Table 1). For eleven sites where the altitude is known, the best fit relationship between GTOPO30 estimates and true altitude is given by altitude estimate51.07#known altitude110.42 m. The correlation is highly significant, R2 = 0.977, n = 11. A thermal lapse rate with altitude of 6(C/km has been used to account for differences in altitude between the cave site and the nearest weather station. Pleistocene thermal histories of the sites have been reconstructed by applying one of four different regional models, (i) northern France and the Benelux countries, (ii) Germany, (iii) Poland (based on data in Aalbersberg & Litt, 1998; Huijzer & Vandenberghe 1998; Coope et al., 1998; Witte et al., 1998) and (iv) Southern France (Fig. 1). The thermal history of southern France has 206 Table 1 Ranking of European Neanderthal Cave sites based upon thermal age Ochoz Khulna Sipka Spy Subalyuk Engis Fond de Foret La Naulette Soulabe las Maretas Neanderthal Saint Brelade Arcy sur Cure La Chaise (B.-Delauney) Malarnaud Rigabe Regourdou Rochelot Montgaudier Caminero Fontechevade Petit Puymoyen Chateauneuf sur C Age kyr Palaeoclimate Model Altitude (m) MAT(C Holocene NA NA NA NA NA NA NA NA NA 50 NA 32 NA NA NA NA NA NA NA NA NA NA P P P NF P NF NF NF SF G NF SF SF SF SF SF SF SF SF SF SF SF 524 514 432 179 416 172 183 168 754 128 10 125 393 475 400 196 141 128 131 120 131 102 6.9 7 7.1 8.9 8.2 9.3 9.5 9.6 8.9 10.1 11.1 11.4 10.4 10.5 11.4 11.6 11.9 12 12.1 12.1 12.1 12.2 Thermal Age (kyr@10(C) 19 21 Thermal Age (kyr@10(C) 10 kyr 32 kyr 50 kyr 5 6 6 8 7 9 9 9 8 10 12 13 11 11 13 14 14 15 15 15 15 15 11 11 11 13 13 14 14 15 13 16 19 21 18 18 21 22 23 24 24 24 24 25 13 13 14 16 16 16 17 17 16 19 21 26 22 22 26 27 29 29 30 30 30 31 at 74 kyr 130 kyr 17 17 17 19 19 20 20 20 21 22 25 34 28 28 34 35 37 38 39 39 39 39 32 33 33 34 38 35 37 37 38 43 47 61 50 51 61 63 67 68 70 70 70 71 C.I. Smith et al. / Journal of Human Evolution 45 (2003) 203–217 Site Table 1 (continued) Site Palaeoclimate Model Altitude (m) MAT(C Holocene NA NA NA NA NA NA NA NA NA NA NA NA NA NA 130 NA NA 32 32 SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF Holocene Holocene, 0(C Pliestocene 390 106 103 355 71 320 169 150 149 108 266 71 75 12 393 280 25 1310 1310 12.2 12.2 12.2 12.3 12.5 12.7 12.7 12.8 12.9 13.1 13.3 13.3 13.3 14.7 10.4 15.2 16.1 5.7 5.7 Thermal Age (kyr@10(C) 50 14 7 Thermal Age (kyr@10(C) 10 kyr 32 kyr 50 kyr 15 15 15 15 16 17 17 17 17 18 19 19 19 24 11 26 31 4 4 25 25 25 25 26 27 27 28 28 29 31 31 31 40 18 44 52 14 7 31 31 31 31 32 34 34 34 35 36 38 38 38 49 22 54 64 N/A N/A at 74 kyr 130 kyr 39 39 39 40 42 43 43 44 45 47 49 49 49 64 28 70 83 N/A N/A 71 71 71 72 75 78 78 80 81 84 88 88 88 115 50 126 149 N/A N/A NB. DNA has been successfully amplified from material from Feldhofer (Krings et al., 1997, Schmitz et al., 2002), the later study dating material to 40 kyr BP in which case the thermal age of the material is w17kyr@10(C and Mezmaiskaya (Ovchinikov et al., 2000); an unsuccessful attempt has been made from material from La Chaise Abri Suard (Cooper et al., 1997). C.I. Smith et al. / Journal of Human Evolution 45 (2003) 203–217 Hortus Marillac Rene Simard Montmaurin Quinzano Orgnac Pech de l’Aze La Masque La Chapelle aux Saints Combe Grenal Lezetxiki Monsempron Roc de Marsal Grotta del Principe La Chaise (Abri Suard) Caverna delle Fate Lazaret Mezmaiskaya (constant temp) Mezmaiskaya (with cooling) Age kyr 207 208 C.I. Smith et al. / Journal of Human Evolution 45 (2003) 203–217 47 Igtocp{ Rqncpf P0"Htcpeg 42 Ogcp"Cppwcn"Vgorgtcvwtg"*³E+ U0"Htcpeg Ncmg"Owpiq 37 U0"Htcpeg"*chvgt"Xcp"Gpfgn+ 32 7 2 42 62 82 :2 322 342 362 /7 /32 Vkog"M[gctu"DR Fig. 1. Regional palaeoclimate models for Northwest Europe and Lake Mungo Australia. Holocene (0–10.4 kyrBP) temperatures are the average mean air temperature taken from the ISLSCP data set (Meeson et al., 1995; Sellers et al., 1995) for Germany, Poland, S. France and N. France. S. France alternative based on Van Andel, 1997, and Lake Mungo on Ambrose, 1984 and Miller et al., 1997. been reconstructed using quantitative data from Guiot et al. (1993), compiled using methods similar to those of the other studies. In cases where palaeoclimatic data are incomplete, palaeotemperature values have been assigned based on similar climatic periods for the same region, or values in similar regions for the same period. The average difference between modern mean annual temperature taken from the International Satellite Land Surface Climatology Project (ISLSCP 1) data set (Meeson et al., 1995; Sellers et al., 1995) and that of the reconstructed mean annual temperature for the last glacial maximum (LGM) has been calculated for each region. To assign each site a temperature for the last glacial maximum, the average LGM/Holocene difference is subtracted from the Holocene temperature for each site. All other differences between Holocene and palaeotemperatures are calculated relative to this LGM/Holocene difference. Thus each site has a palaeotemperature curve based on its current (Holocene) temperature, and its regional palaeoclimatic curve. There is an implicit assumption in this approach; that current differences in temperature caused by geographic location etc. are constant with time. In part this assumption is true, as the main differences in temperature between sites (especially within one region) will be due to differences in altitude, these differences are maintained in this model (although this does assume that the thermal lapse rate remains constant). Conversely the palaeoclimatic evidence suggests that the temperature gradient across Europe, currently northsouth, has fluctuated in the past, in both direction C.I. Smith et al. / Journal of Human Evolution 45 (2003) 203–217 and inclination, and local effects such as vegetation cover and local weather conditions will have changed. These factors will affect both inter- and intra-regional variation, and are not accounted for in the model. The reconstructions also ignore the possibility that geothermal heating may have elevated cave temperatures during periods of reduced precipitation in the Pleistocene. An alternative climate model for southern France has been used, for comparative purposes, based on the climate curve produced by Van Andel (1997), essentially a reconstruction based on palaeoclimatic data from Les Echets (Guiot et al., 1989). This model is not confined to the chronology of the data above, but only covers the last 60 kyr. This model has been used to compare how thermal ages differ using alternative palaeoclimate models. During periods of glaciation it is likely that cave deposits underwent periods of freezing. The relationship between air temperature and permafrost surface temperature is difficult to predict due to the zero curtain effect (Zhang et al., 1997). However, the rate of depurination in frozen samples is thought to be weakly temperature sensitive (Osborne & Phillips, 2000) so this problem is somewhat circumvented. For the purposes of modelling we have taken a nominal effective temperature of 0(C for fossils in permafrost, i.e. where the model predicts a temperature of 0(C or less. The thermal age of Mezmaiskaya has also been reconstructed, although we lack quantitative palaeoclimatic data for this region. We have therefore used two reconstructions for comparison: one using a constant Holocene temperature throughout, the other using 0(C for the first 22 kyr of deposition and the Holocene temperature for the last 10 kyr. Thermal model for Lake Mungo The thermal history of Lake Mungo has been reconstructed using measured sediment temperatures from Lake Mungo (Ambrose, 1984) and reconstructed palaeotemperatures inferred from isoleucine epimerisation rates from ratite eggshells (Miller et al., 1997). The data for burials at 1.5 m at stations 1 and 2 (Ambrose, 1984) were used to calculate the effective burial temperatures. 209 Effective DNA depurination temperature (Teff) is the constant temperature equivalent which causes the net equivalent depurination that will have been accrued by the DNA over one year, taking into account seasonal temperature fluctuation and using a temperature dependent reaction having an activation energy of Ea 127 kJ mol
1 (Lindahl & Nyberg, 1972). Total annual reaction was calculated using a sinusoidal wave model using the mean annual temperature and variation about this mean for modern temperatures and palaeotemperatures (see Appendix B for an example of the calculation of thermal age). The palaeoclimate of Southern Australia is markedly different to that of Northwest Europe. A model based upon amino acid epimerisation data indicates a 9(C warming after 16 kyr BP with no other significant excursions (Miller et al., 1997). We have therefore used a simple palaeoclimatic model using the present day effective temperatures calculated from sediment temperatures for the period 16-0 kyr BP, which was preceded by a period 9(C cooler than the present arithmetic mean annual temperature, from 60 kyr BP-16 kyr BP (Fig. 1). In addition we have calculated the thermal age at constant mean temperature, to account for dampening effects assuming that the burial was at great depth for most of its history. Rates of biomolecular deterioration The amount of DNA depurination at each site has been calculated as the integration of the site’s palaeoclimatic curve, the age of the fossils at the site, and the activation energy (Ea) of the depurination reaction (Lindahl & Nyberg, 1972). To enable comparison, sites are ranked using ‘thermal age’, where the calculated rates are normalised to what the rate would be at 10(C (see Appendix B). Sites that have multiple occupancy or have not been accurately dated, a number of alternative thermal ages are given corresponding to different actual periods of burial. Results and discussion The results of this analysis can be seen in Table 1. The Feldhofer cave site from which DNA 210 C.I. Smith et al. / Journal of Human Evolution 45 (2003) 203–217 has been amplified is ranked as ten of thirty-nine, with a thermal age of w19kyr@10(C. If our estimate of the Holocene temperature is 1(C in error at this site, the thermal age is approximately 3kyr@10(C (w16%) different. The sequence of ages given for each site reveals an important feature of the thermal history. For example, at Feldhofer, if a fossil is considered to be 50 kyr, 45% of the thermal age is accumulated over the first 40 kyr of deposition, and 55% of the thermal age is accumulated during the last 10 kyr. Recent radiocarbon determinations have directly re-dated the fossil humans at Feldhofer to about 40 kyr (Schmitz et al., 2002), this re-dating to 40 kyr will reduce the thermal age by only 2kyr@10(C. Furthermore, if a fossil was deposited 74 kyr BP at Feldhofer, the thermal age would only be w3kyr@10(C older than a fossil deposited at 50 kyr BP. This reveals the importance of considering the entire thermal history of a fossil, and the effect of the warming during the Holocene. This also indicates that the temperature at which the fossil is held during the Holocene is of key importance to biomolecular survival, and that the absolute age of the fossil may be of negligible importance if the fossil is kept very cold. Using the alternative climate model based on Van Andel (1997) a southern French site with a current estimated temperature of 12.2(C (e.g. Marillac) will have its thermal age reduced by approximately 1kyr@10(C at 50 kyr, and by 2kyr@10(C at 74 kyr (3–5%). This demonstrates some of the error involved in estimating thermal age, and highlights the importance of the climate model. The thermal age of Mezmaiskaya (also a site where aDNA has been successfully amplified; Ovchinikov et al., 2000) is likely to have been overestimated here, where we have not modelled any cooling during the Pleistocene, but it is still lower than that of Feldhofer cave. This is probably due to the altitude of the site (1.3 km), keeping the site cold in comparison with others. If published data of aDNA amplification success and failure is compared with the age of the material, there is no apparent pattern of survival, however, if it is compared to the thermal age of the material, then it becomes apparent that only thermally young material is likely to yield amplifiable aDNA (Fig. 2). Most successes come from either permafrost deposits or young material (<5 kyrs) from temperate regions (mainly Europe). Thus, one might expect the limit of DNA survival to be defined by the thermally oldest material from this area, i.e. Feldfoher (Krings et al., 1997) and Scladina (Lorreille et al., 2001), particularly when considering the technical difficulties of these studies. When the thermal age of Lake Mungo is compared to other studies, the thermal age is far greater than any other reported successes. Pääbo & Wilson (1991) used a theoretical approach, based upon the rate of depurination of DNA at 15(C and pH of 7.0 in physiological solution (Lindahl & Nyberg, 1972), to suggest that it would take approximately 5 ka to destroy the last amplifiable 800 bp of chloroplast DNA (cpDNA) fragment in 1 g of leaf tissue given a starting content of 1012 cpDNA g
1. Marota et al. (2002) estimated the survival limit of cpDNA in papyri at 35(C to be w800 years, using a reaction rate for depurination at 15(C based upon the values calculated by Pääbo & Wilson (1991), but instead of using the published activation energy (127 kJ m
1, Lindahl & Nyberg, 1972) supposed a trebling of the rate of reaction per 10(C temperature increase (i.e. 76 kJ m
1). We have used a slightly modified approach to calculate the survival of the last fragment, using the original Lindahl & Nyberg (1972) kinetics (see Appendices). For the original conditions described by Pääbo & Wilson (1991) our estimate for the median survival time of the last surviving 800 bp fragment, from an original 1012 copies is an order of magnitude lower, 359 years (and the first and 99th percentiles of survival time are estimated at 334 and 413 years respectively). The corresponding estimate for 35(C (Marota et al., 2002), is 11.5 years (10.7 and 13.2; 1st and 99th percentiles, respectively). Assuming the copy number of mitochondrial DNA (mtDNA) in fresh bone to be about 2.51011 g
1 (Table 2) we estimate the survival of a 105 bp fragment in 0.4g of bone (the fragment and sample size used by Krings et al., 1997, to successfully amplify aDNA from the Neanderthal type specimen) at 10(C and pH 7.4 to be 15 ka (14 ka and 17 ka; 1st and 99th percentiles), extending to w107 ka at 0(C (100 ka and 124 ka, 1st and 99th percentiles). 211 C.I. Smith et al. / Journal of Human Evolution 45 (2003) 203–217 382 Ncmg"Owpiq 362 Vjgtocn"Cig"*m{tuB32³E+ 342 322 :2 82 62 Hgnfjqhgt 42 2 42 62 82 :2 322 342 Cig"*m{tu+ Fig. 2. Comparison of thermal age (@10(C) and fossil age for reports of DNA amplification success and failure. Full squares represent successful DNA amplification and open circles failures. Feldhofer is represented as a range reflecting the uncertainty of the age of the site. The range of Thermal Age at lake Mungo represents uncertainty in the thermal model to use due to possible different burial histories. (Data taken from Adcock et al., 2001; Bailey et al., 1996; Colson et al., 1997; Cooper et al., 1996; Cooper et al., 1997; Fleischer et al., 2000; Hoss et al., 1994; Krings et al., 1997; Leonard et al., 2000; Leonard et al., 2002; Loreille et al., 2001; Ovchinikov et al., 2000; Stone & Stoneking, 1999; Vilà et al., 2001). Our thermal age limit of w19kyr@10(C is thus beyond what may be estimated from the copy number calculation, 15 kyr. Copy number estimates of 105 bp fragments from the Feldhofer bones however, are approximately 2.5–3.75103 copies remaining per gram of bone (Krings et al., 1997). Using the same parameters, this would suggest that the rate would need to be approximately 4.9 times slower than the rate at pH 7.4 at 10(C. The rate of depurination at pH 7.88 is approximately twice as slow as that at pH 7.4, and is a more realistic pH for reactions in bone. Furthermore, Lindahl (1993) reports a two-fold reduction in depurination rate when adsorbed to apatite. Marguet & Forterre (1998) report that rates of depurination are retarded for both double stranded and single stranded DNA in the presence of chloride salts, due to their direct interaction with purine nucleotides; thus elevated salt concentrations may also reduce the predicted rates of DNA deterioration. Differences in absolute rates will not affect the relative rankings of the sites given in Table 1, but would compromise predictions of the remaining copy numbers (c.f. Pääbo & Wilson, 1991); absolute rates of DNA deterioration are therefore not estimated for the Neanderthal sites. Given the technical difficulty of amplifications from the Neanderthal material and the variable state of bone preservation, we would suggest that in the absence of convincing screening methods, (see Collins et al., 1999) 19kyr@10(C is a sensible thermal limit. Note, however, that this does not mean that all sites with thermal ages lower than 212 value 1 Volume occupied by osteocyte lacunae 2 Volume occupied by osteocyte lacunae 3 Size of individual osteocyte lacunae µm 4 5 6 7 8 9 10 11 Hence volume of osteocyte Volume of osteocyte occupied by mitochondria Skeletal Density of modern bone gcm
3 Hence numbers of osteocytes per gram modern bone No of mitochondria per cell Hence no of copies of mitochondria per gram No of copies of mtDNA per mitochondria No of copies of mtDNA per gram of bone HgIP, Mercury Intrusion Porosimetry. 1.5% of bone 1.7–2.8% 11.7–17.4 length 4.8–6.6 width 3–3.4 height 200 µm3 20% 1.4–1.7 5107 1000 51010 5 2.51011 Comment HgIP of human bone (NB slightly higher values are seen in bird bone) http://www.orl.med.umich.edu/orl/archgroup/a6.htm http://www.orl.med.umich.edu/orl/archgroup/a6.htm in liver cells values of 15–25% reported increases ultimately to 3.0 due to loss of collagen and water assuming a density of 1.47 gcm
3 Ranges from 800–1500 depending upon values given in (5) literature values range from 2–20 estimate from 31010-11012 using the extremes given in (5) and (10) C.I. Smith et al. / Journal of Human Evolution 45 (2003) 203–217 Table 2 Estimation of mtDNA content in compact bone 213 C.I. Smith et al. / Journal of Human Evolution 45 (2003) 203–217 Table 3 Modelled thermal ages of Lake Mungo fossils if buried at 1.5 m Station Year Mean Soil at 1.5 m depth Temperature* Soil Temperature at 1.5 m depth Amplitude* Effective temperature after 16 kyr BP Effective Temperature before 16 kyr BP Thermal age (kyr@10C) if 60 kyr Thermal age (kyr@10C) if 40 kyr Thermal age (kyr@10C) if 60 kyr (no amplitude) Thermal age (kyr@10C) if 40 kyr (no amplitude) 1 82/83 20.1 2.5 20.4 11.1 162 137 157 132 1 83/84 19.1 1.7 19.2 10.1 133 112 131 110 2 82/83 20.6 7 22.5 11.6 216 189 172 145 2 83/84 19.8 5.9 21.2 10.8 176 152 148 125 *Data taken from Ambrose (1984). Feldhofer will contain amplifiable aDNA, as other factors may accelerate deterioration (Ovchinnikov et al., 2001; Collins et al., 2002). Furthermore, by amplification of shorter fragments it is possible to extend the temporal range further. Thus reports of the successful amplification of an 88 bp product from a cave bear 80 kyr–100 kyr at Scladina Cave (Loreille et al., 2001) with a thermal age of w20kyr@10(C are consistent with the results from Feldhofer (considering the actual rate of DNA degradation at Feldhofer is slower than we predict). Using data from Ambrose (1984), a number of thermal models for the Lake Mungo site can be established (Table 3) based upon differences in temperature recorded at different times and different depths. Although the range of calculated ‘thermal ages’ for a 60 kyr fossil is large (131– 216kyr@10(C), they are an order of magnitude higher than that of Feldhofer cave. If Feldhofer cave lies close to the technical limit of amplification at 19kyr@10(C it would seem highly unlikely that there would be amplifiable DNA in the Lake Mungo fossils (especially as the longer fragment lengths of 153 to 189 bp targeted by Adcock et al. (2001) will decay 1.4 and 1.7 times more rapidly than a comparable 105 bp product). If the site is only 40 kyr old as has been suggested (Gillespie & Roberts, 2000; Bowler et al., 2003; but compare Grün et al., 2000), the thermal age is reduced by at most 16%, still much higher than that of Feldhofer. The extreme thermal age at the site, the poor state of preservation of the material, acknowledged as being too fragmentary to sex reliably (Brown, 2000; Thorne & Curnoe, 2000), and having “negligible organic preservation’ (Gillespie & Roberts, 2000 p. 727) indicate that DNA preservation at the site as a whole is likely to be poor. This coupled with the concerns raised over the technical aspects of the Lake Mungo sequences (Cooper et al., 2001) suggest that the reports of amplification of aDNA from Lake Mungo should be treated with caution until the work can be better substantiated, e.g., repeated in another laboratory. If DNA can be extracted from fossils with thermal ages as high as that of Lake Mungo the potential for finding aDNA in other samples, particularly many of the Neanderthal sites, are greatly increased. Conclusion The application of molecular techniques to fossil materials is an expanding field, and can provide invaluable data on the relationships between modern humans and their fossil relatives. At present only a few authentic sequences of fossil hominid DNA have been reported. Without a better understanding of the survival of DNA in the fossil record, valuable fossils may be damaged and destroyed and much time wasted in the search for more aDNA sequences. The thermal history of fossils is a useful indicator for the likelihood of survival of biomolecules in the fossil record. If DNA depurination is considered to be the key mechanism of DNA degradation in fossil bone material (excluding the role of micro-organisms, 214 C.I. Smith et al. / Journal of Human Evolution 45 (2003) 203–217 which may be a prerequisite for long term survival, Trueman & Martill, 2002), then the thermal history of the site will be a key parameter in defining the state of DNA preservation. The analysis of the thermal history of northwest European cave sites from which Neanderthal remains have been recovered, and subsequent conversion to a relative amount of DNA degradation, indicates that few other northwest European Neanderthal cave sites are likely to yield amplifiable DNA using present techniques. Furthermore when the relative amount of DNA degradation at Lake Mungo is compared with that of northwest European cave sites, we would predict that the likelihood of amplifying authentic endogenous DNA from fossils from this site is very low. If this is true, then the authenticity of this sequence must be brought further into doubt (Cooper et al., 2001). The pth percentile, tp, of the distribution of T0 is, therefore, given by 1 N F S DG 1 p 1n 1⫺ tp⫽⫺ lbpk 100 Appendix B: Calculation of DNA thermal age The thermal age of a fossil is calculated as in the following example, for a 16 kyrBP fossil bone from Lake Mungo (Station 1, 1.5 m depth), using temperature data for the year 1982/1983 (Ambrose 1984). The modern day effective temperature Teff is calculated using a simple sinusoidal model of soil temperature variation throughout the year, based upon soil temperatures, with mean 20.1(C and amplitude 2.5(C Changes in rate of reaction are calculated using the formula Acknowledgements k⫽Ae⫺Ea/RT The LDEO/IRI Data Library is acknowledged for facilitating access to climatic data. Jen Hiller is thanked for allowing us access to her temperature data from Scladina cave and Marcel Otte and Dominique Bonjean for allowing access to this site. The rate of depurination (k yr
1) in bone is estimated at pH 7.4 from data given in Lindahl & Nyberg (1972), an activation energy, Ea = 127 kJ and a pre-exponential constant, mol
1 A = 1.451011 s
1, which is slightly lower than that used by Pääbo & Wilson (1991; A = 3.261011 s
1 at pH 7.0), have been used here, and R = 8.314 Jmol
1 K
1 and T is the temperature (in Kelvin) of the reaction. As the rate of rate of depurination is exponentially related to the reciprocal of temperature the amplitude of the sine wave is an important factor. The average rate of depurination is calculated for 1 year (kavg) accounting for the exponential increase in rate. This rate of reaction is approximately equivalent to the sample being held at a constant 20.4(C, the effective DNA depurination temperature. This procedure is reproduced for each temperature in the thermal history of a fossil. The relative amount of DNA depurination at each site has been normalised to the equivalent amount of damage to the DNA as if the sample were held at a constant 10(C, and reported as the thermal age. Therefore, in the above example, for the last 16 kyr of burial the rate of reaction at Appendices Appendix A: Calculation of the probability of the last amplifiable fragment remaining with time The calculation is based upon the assumptions that random scission occurs and that all fragments are equally vulnerable to degradation. Let T0 be the time at which the last amplifiable fragment is lost. Then the probability that the loss of the last amplifiable fragment occurs at or before time t is given by P(T0#t)⫽[1⫺exp(⫺lbpkt)]N where N is the initial copy number lbp is the length of fragment in base pairs k is the rate of reaction C.I. Smith et al. / Journal of Human Evolution 45 (2003) 203–217 20.4(C is approximately 6.71 times faster than that at 10(C, i.e., every year at this temperature is equivalent to 6.71 years at 10(C. For this period the sample accumulates 16 kyr6.71 thermal years = 107kyr@10(C. References Aalbersberg, G., Litt, T., 1998. Multiproxy climate reconstructions for the Eemian and Early Weichselian. Journal of Quaternary Science 13, 367–390. Adcock, G.J., Dennis, E.S., Easteal, S., Huttley, G.A., Jermiin, L.S., Peacock, W.J., Thorne, A., 2001. Mitochondrial DNA sequences in ancient Australians: Implications for modern human origins. Proceedings of the National Academy of Science, USA 98, 537–542. Ambrose, W.R., 1984. Soil Temperature Monitoring at Lake Mungo: Implications for racemisation dating. Australian Archaeology 19, 64–74. Bada, J.L., Shou, M.Y., 1980. Kinetics and mechanisms of amino acid racemization in aqueous solution and bones. In: Hare, P.E., Hoering, T.C., King, K.J. (Eds.), Biogeochemistry of amino acids. Wiley, New York, pp. 235–255. Bailey, J.F., Richards, M., Macaulay, V.A., Colson, I.B., James, I.T., Bradley, D.G., Hedges, R.E.M., Sykes, B.C., 1996. Ancient DNA suggests a recent expansion of European cattle from a diverse wild progenitor species. Proceedings of the Royal Society of London B 263, 1467–1473. Baker, C.B., Eischeid, J.K., Karl, T.R. & Diaz, H.F. (1994). The quality control of long-term climatological data using objective data analysis. Preprints of AMS Ninth Conference on Applied Climatology, Dallas, TX. January, 15–20 1995. Bell, L.S., Skinner, M.F., Jones, S.J., 1996. The speed of post mortem change to the human skeleton and its taphonomic significance. Forensic Science International 82, 129–140. Bogli, A., 1980. Karst Hydrology and Physical Speleology. (Translated by J.C. Schmid). Springer Verlag, New York. Bowler, J.M., Johnston, H., Olley, J.M., Prescott, J.R., Roberts, R.G., Shawcross, W., Spooner, N.A., 2003. New ages for human occupation and climatic change at Lake Mungo, Australia. Nature 421, 837–840. Brown, P., 2000. Australian Pleistocene variation and the sex of Lake Mungo 3. Journal of Human Evolution 38, 743–749. Collins, M.J., Waite, E.R., van Duin, A.C.T., 1999. Predicting protein decomposition, the case of aspartic acid racemization kinetics. Philosophical Transactions of the Royal Society, Series B 354, 51–64. Collins, M.J., Nielsen-Marsh, C.M., Hiller, J., Smith, C.I., Roberts, J.P., Prigodich, R.V., Wess, T.J., Csapo, J., Millard, A., Turner-Walker, G., 2002. The survival of organic matter in bone: a review. Archaeometry 44, 383–394. 215 Colson, I.B., Bailey, J.F., Vercauteren, M., Sykes, B.C., 1997. The Preservation of Ancient DNA and Bone Diagenesis. Ancient Biomolecules 1, 109–117. Coope, G.R., Lemdahl, G., Lowe, J.J., Walking, A., 1998. Temperature gradients in northern Europe during the last glacial Holocene transition (14-9 14C kyr BP) interpreted from coleopteran assemblages. Journal of Quaternary Science 13, 419–433. Cooper, A., Poinar, H.N., 2000. Ancient DNA: Do It Right or Not at All. Science 289, 1139–1139. Cooper, A., Poinar, H., Pääbo, S., Radovcic, J., Debénath, A., Caparros, M., Barroso-Ruiz, C., Bertranpetit, J., Nielsen-Marsh, C., Hedges, R.M., Sykes, B., Clark, G.A., 1997. Neandertal Genetics. Science 277, 1022–1025. Cooper, A., Rambaut, A., Macaulay, V., Willerslev, E., Hansen, A.J., Stringer, C., 2001. Human origins and ancient human DNA. Science 292, 1655–1656. Cooper, A., Rhymer, J.D., James, H.F., Olson, S.L., McIntosh, C.E., Sorenson, M.D., Fleischer, R.C., 1996. Ancient DNA and island endemics. Nature 381, 484. Fleischer, C.F., Olson, S.L., James, H.F., Cooper, A., 2000. Identification of the extinct Hawaiian Eagle (Haliaeetus) by mtDNA sequence analysis. Auk 117, 1051–1056. Gesch, D.B., Verdin, K.L., Greenlee, S.K., 1999. New Land Surface Digital Elevation Model Covers the Earth. EOSM Transactions, AGU 80, 69–70. Gillespie, R., Roberts, R.G., 2000. On the reliability of age estimates for human remains at Lake Mungo. Journal of Human Evolution 38, 727–732. Grün, R., Spooner, N.A., Thorne, A., Mortimer, G., Simpson, J.J., McCulloch, M.T., Taylor, L., Curnoe, D., 2000. Age of the Lake Mungo 3 skeleton, reply to Bowler & Magee and to Gillespie & Roberts. Journal of Human Evolution 38, 733–741. Guiot, J., Pons, A., deBeaulieu, J.L., Reille, M., 1989. A 140,000-year continental climate reconstruction from two European pollen records. Nature 338, 309–313. Guiot, J., de Beaulieu, J.L., Cheddadi, R., David, F., Ponel, P., Reille, M., 1993. The climate in western Europe during the last glacial/interglacial cycle derived from pollen and insect remains. Palaeogeography, Palaeoclimatology, Palaeoecology 103, 73–93. Hackett, C.J., 1981. Microscopical focal destruction (tunnels) in exhumed human bones. Medicine, Science, Law 21, 243–265. Hedges, R.E.M., Millard, A.R., 1995. Bones and groundwater: towards the modelling of diagenetic processes. Journal of Archaeological Science 22, 155–164. Höss, M., 2000. Neanderthal population genetics. Nature 404, 453–454. Höss, M., Pääbo, S., Vereshchagin, N.K., 1994. Mammoth DNA sequences. Nature 370, 333. Huijzer, B., Vandenberghe, J., 1998. Climate reconstruction of the Weichselian Pleniglacial in northwestern and central Europe. Journal of Quaternary Science 13, 391–417. Krarti, M., Lopez-Alonzo, C., Claridge, D.E., Kreider, J.F., 1995. Analytical Model to Predict Annual Soil Surface 216 C.I. Smith et al. / Journal of Human Evolution 45 (2003) 203–217 Temperature Variation. Transactions of the ASME 117, 91–99. Krings, M., Stone, A., Schmitz, R.W., Krainitzki, H., Stoneking, M., Pääbo, S., 1997. Neanderthal DNA Sequences and the Origin of Modern Humans. Cell 90, 19–30. Krings, M., Capelli, C., Tschentscher, F., Geisert, H., Meyer, S., von Haeseler, A., Grossschmidt, K., Possnert, G., Paunovic, M., Pääbo, S., 2000. A view of Neanderthal genetic diversity. Nature Genetics 26, 144–146. Kusada, T., Achenbach, P.R., 1965. Earth Temperature and Thermal Diffusivity at Selected Stations in the United States. ASHRAE Transactions 71, 61–75. Leonard, J.A., Wayne, R.K., Cooper, A., 2000. Population Genetics of Ice Age Brown Bears. Proceeding of the National Academy of Sciences, USA 97, 1651–1654. Leonard, J.A., Wayne, R.K., Wheeler, J., Valadez, R., Guillen, S., Vilà, C., 2002. Ancient DNA Evidence for Old World Origin of New World Dogs. Science 298, 1613–1616. Lindahl, T., 1993. Instability and decay of the primary structure of DNA. Nature 362, 709–715. Lindahl, T., Nyberg, B., 1972. Rate of depurination of native deoxyribonucleic acid. Biochemistry 11, 3610–3618. Loreille, O., Orlando, L., Patou-Mathis, M., Philippe, M., Taberlet, P., Hänni, C., 2001. Ancient DNA analysis reveals ancient divergence of the cave bear, Ursus spelaeus, and brown bear, Ursus arctos, lineages. Current Biology 11, 200–203. Marguet, E., Forterre, P., 1998. Protection of DNA by salts against degradation at temperatures typical for hyperthermophiles. Extremophiles 2, 115–122. Marota, I., Basile, C., Ubaldi, M., Rollo, F., 2002. DNA decay rate in papyri and human remains from Egyptian archaeological sites. American Journal of Physical Anthropology 117, 310–318. Meeson, B.W., Corprew, F.E., McManus, J.M.P., Myers, D.M., Closs, J.W., Sun, K-J., Sunday, D.J. & Sellers P.J. (1995). ISLSCP Initiative I - Global Data Sets for LandAtmosphere Models, 1987–1988. Volumes 1–5. (CD NASA, 1995) Mellars, P., 1999. The Neanderthal Problem Continued. Current Anthropology 40, 341–364. Miller, G.H., Magee, J.W., Jull, A.J.T., 1997. Low-latitude cooling in the Southern Hemisphere from amino-acid racemization in emu eggshells. Nature 385, 241–244. Osborne, M.R., Phillips, D.H., 2000. Preparation of a Methylated DNA Standard, and Its Stability on Storage. Chem. Res. Toxicol. 13, 257–261. Ovchinnikov, I., Götherström, A., Romanova, G.P., Kharitonov, V.M., Liden, K., Goodwin, W., 2000. Molecular analysis of Neanderthal DNA from the northern Caucasus. Nature 404, 490–493. Ovchinnikov, I., Götherström, A., Romanova, G.P., Kharitonov, V.M., Liden, K., Goodwin, W., 2001. Neanderthal DNA; not just old but old and cold - reply to Smith et al. Nature 410, 772. Pääbo, S., Wilson, A.C., 1991. Miocene DNA sequences-a dream come true? Current Biology 1, 45–46. Pischetsrieder, M., Seidel, W., Munch, G., Schinzel, R., 1999. N-2-(1-carboxyethyl)deoxyguanosine, a nonenzymatic glycation adduct of DNA, induces single-strand breaks and increases mutation frequencies. Biochemical and Biophysical Research Communications 264, 544–549. Poinar, H.N., 1999. DNA from fossils: the past and the future. Acta Paediatrica 88, 133–140. (Supplement). Schmitz, R.W., Serre, D., Bonani, G., Feine, S., Hillgruber, F., Krainitzki, H., Pääbo, S., Smith, F.H., 2002. The Neanderthal type site revisited: Interdisciplinary investigations of skeletal remains from the Neander Valley, Germany. Proceedings of the National Academy of Science, USA 99, 13342–13347. Sellers, P.J., Meeson, B.W., Closs, J., Collatz, J., Corprew, F., Dazlich, D., Hall, F.G., Kerr, Y., Koster, R., Los, S., Mitchell, K., McManus, J., Myers, D., Sun, K-J. & Try, P. (1995). An overview of the ISLSCP Initiative I Global Data Sets. On: ISLSCP Initiative I Global Data Sets for LandAtmosphere Models, 1987–1988. Volumes 1–5. (CD NASA, 1995). Smith, C.I., Chamberlain, A.T., Riley, M.S., Cooper, A., Stringer, C.B., Collins, M.J., 2001. Neanderthal DNA; not just old but old and cold. Nature 410, 771–772. Stone, A.C., Stoneking, M., 1999. Analysis of ancient DNA from a prehistoric Amerindian cemetery. Philosophical Transactions of the Royal Society of London Series B 354, 153–159. Thorne, A., Curnoe, D., 2000. Sex and significance of Lake Mungo 3: reply to Brown “Australian Pleistocene variation and the sex of Lake Mungo 3”. Journal of Human Evolution 39, 587–600. Toy, T.J., Kuhaida, J.N.R.A.J., Munson, B.E., 1978. The prediction of mean monthly soil temperature from mean monthly air temperature. Soil Science 126, 181–189. Trueman, C.N., Martill, D.M., 2002. The long-term survival of bone: the role of bioerosion. Archaeometry 44, 371–382. Turner-Walker, G, Nielsen-Marsh, C.M., Syversen, U., Kars, H., Collins, M.J., (in press) Sub-micron spongiform porosity is the major ultra-structural alteration occurring in archaeological bone. International Journal of Osteoarchaeology. Van Andel, T.H., 1997. Middle and Upper Palaeolithic environments and the calibration of 14C dates beyond 10,000 BP. Antiquity 72, 26–33. Vilà, C., Leonard, J.A., Götherström, A., Marklund, S., Sandberg, K., Lidén, K., Wayne, R.K., Ellegren, H., 2001. Widespread origins of domestic horse lineages. Science 291, 474–477. Wayne, R.K., Leonard, J.A., Cooper, A., 1999. Full of Sound and Fury: History of Ancient DNA. Annu. Rev. Ecol. Syst. 30, 457–477. Wehmiller, J.F., Stecher, H.A. III, York, L., Friedman, I., 2000. The Thermal Environment of Fossils: Effective Ground Temperatures (1994–1998) at Aminostratigraphic Sites, U.S. Atlantic Coastal Plain. In: Goodfriend, G.A., C.I. Smith et al. / Journal of Human Evolution 45 (2003) 203–217 Collins, M.J., Fogel, M.L., Macko, S.A., Wehmiller, J.F. (Eds.), Perspectives in Amino Acid and Protein Geochemistry. Oxford University Press, New York, pp. 219–250. Witte, H.J.L., Coope, G.R., Lemdahl, G., Lowe, J.J., 1998. Regression coefficients of thermal gradients in northwestern Europe during the last glacial-Holocene transition using 217 beetle MCR data. Journal of Quaternary Science 13, 435–445. Zhang, T., Osterkamp, T.E., Stamnes, K., 1997. Effects of climate on the active layer and permafrost on the north slope of Alaska, USA. Permafrost and Periglacial Processes 8, 45–67.